distribution.java

一个数据挖掘系统的源码

      totaL = totaL + weight;
    }
  }

  /**
   * Checks if at least two bags contain a minimum number of instances.
   */
  public final boolean check(double minNoObj) {

    int counter = 0;
    int i;

    for (i=0;i<m_perBag.length;i++)
      if (Utils.grOrEq(m_perBag[i],minNoObj))
	counter++;
    if (counter > 1)
      return true;
    else
      return false;
  }

  /**
   * Clones distribution (Deep copy of distribution).
   */
  public final Object clone() {

    int i,j;

    Distribution newDistribution = new Distribution (m_perBag.length,
						     m_perClass.length);
    for (i=0;i<m_perBag.length;i++) {
      newDistribution.m_perBag[i] = m_perBag[i];
      for (j=0;j<m_perClass.length;j++)
	newDistribution.m_perClassPerBag[i][j] = m_perClassPerBag[i][j];
    }
    for (j=0;j<m_perClass.length;j++)
      newDistribution.m_perClass[j] = m_perClass[j];
    newDistribution.totaL = totaL;

    return newDistribution;
  }

  /**
   * Deletes given instance from given bag.
   *
   * @exception Exception if something goes wrong
   */
  public final void del(int bagIndex,Instance instance)
       throws Exception {

    int classIndex;
    double weight;

    classIndex = (int)instance.classValue();
    weight = instance.weight();
    m_perClassPerBag[bagIndex][classIndex] =
      m_perClassPerBag[bagIndex][classIndex]-weight;
    m_perBag[bagIndex] = m_perBag[bagIndex]-weight;
    m_perClass[classIndex] = m_perClass[classIndex]-weight;
    totaL = totaL-weight;
  }

  /**
   * Deletes all instances in given range from given bag.
   *
   * @exception Exception if something goes wrong
   */
  public final void delRange(int bagIndex,Instances source,
			     int startIndex, int lastPlusOne)
       throws Exception {

    double sumOfWeights = 0;
    int classIndex;
    Instance instance;
    int i;

    for (i = startIndex; i < lastPlusOne; i++) {
      instance = (Instance) source.instance(i);
      classIndex = (int)instance.classValue();
      sumOfWeights = sumOfWeights+instance.weight();
      m_perClassPerBag[bagIndex][classIndex] -= instance.weight();
      m_perClass[classIndex] -= instance.weight();
    }
    m_perBag[bagIndex] -= sumOfWeights;
    totaL -= sumOfWeights;
  }

  /**
   * Prints distribution.
   */

  public final String dumpDistribution() {

    StringBuffer text;
    int i,j;

    text = new StringBuffer();
    for (i=0;i<m_perBag.length;i++) {
      text.append("Bag num "+i+"\n");
      for (j=0;j<m_perClass.length;j++)
	text.append("Class num "+j+" "+m_perClassPerBag[i][j]+"\n");
    }
    return text.toString();
  }

  /**
   * Sets all counts to zero.
   */
  public final void initialize() {

    for (int i = 0; i < m_perClass.length; i++)
      m_perClass[i] = 0;
    for (int i = 0; i < m_perBag.length; i++)
      m_perBag[i] = 0;
    for (int i = 0; i < m_perBag.length; i++)
      for (int j = 0; j < m_perClass.length; j++)
	m_perClassPerBag[i][j] = 0;
    totaL = 0;
  }

  /**
   * Returns matrix with distribution of class values.
   */
  public final double[][] matrix() {

    return m_perClassPerBag;
  }

  /**
   * Returns index of bag containing maximum number of instances.
   */
  public final int maxBag() {

    double max;
    int maxIndex;
    int i;

    max = 0;
    maxIndex = -1;
    for (i=0;i<m_perBag.length;i++)
      if (Utils.grOrEq(m_perBag[i],max)) {
	max = m_perBag[i];
	maxIndex = i;
      }
    return maxIndex;
  }

  /**
   * Returns class with highest frequency over all bags.
   */
  public final int maxClass() {

    double maxCount = 0;
    int maxIndex = 0;
    int i;

    for (i=0;i<m_perClass.length;i++)
      if (Utils.gr(m_perClass[i],maxCount)) {
	maxCount = m_perClass[i];
	maxIndex = i;
      }

    return maxIndex;
  }

  /**
   * Returns class with highest frequency for given bag.
   */
  public final int maxClass(int index) {

    double maxCount = 0;
    int maxIndex = 0;
    int i;

    if (Utils.gr(m_perBag[index],0)) {
      for (i=0;i<m_perClass.length;i++)
	if (Utils.gr(m_perClassPerBag[index][i],maxCount)) {
	  maxCount = m_perClassPerBag[index][i];
	  maxIndex = i;
	}
      return maxIndex;
    }else
      return maxClass();
  }

  /**
   * Returns number of bags.
   */
  public final int numBags() {

    return m_perBag.length;
  }

  /**
   * Returns number of classes.
   */
  public final int numClasses() {

    return m_perClass.length;
  }

  /**
   * Returns perClass(maxClass()).
   */
  public final double numCorrect() {

    return m_perClass[maxClass()];
  }

  /**
   * Returns perClassPerBag(index,maxClass(index)).
   */
  public final double numCorrect(int index) {

    return m_perClassPerBag[index][maxClass(index)];
  }

  /**
   * Returns total-numCorrect().
   */
  public final double numIncorrect() {

    return totaL-numCorrect();
  }

  /**
   * Returns perBag(index)-numCorrect(index).
   */
  public final double numIncorrect(int index) {

    return m_perBag[index]-numCorrect(index);
  }

  /**
   * Returns number of (possibly fractional) instances of given class in
   * given bag.
   */
  public final double perClassPerBag(int bagIndex, int classIndex) {

    return m_perClassPerBag[bagIndex][classIndex];
  }

  /**
   * Returns number of (possibly fractional) instances in given bag.
   */
  public final double perBag(int bagIndex) {

    return m_perBag[bagIndex];
  }

  /**
   * Returns number of (possibly fractional) instances of given class.
   */
  public final double perClass(int classIndex) {

    return m_perClass[classIndex];
  }

  /**
   * Returns relative frequency of class over all bags with
   * Laplace correction.
   */
  public final double laplaceProb(int classIndex) {

    return (m_perClass[classIndex] + 1) /
      (totaL + (double) actualNumClasses());
  }

  /**
   * Returns relative frequency of class for given bag.
   */
  public final double laplaceProb(int classIndex, int intIndex) {

    return (m_perClassPerBag[intIndex][classIndex] + 1.0) /
      (m_perBag[intIndex] + (double) actualNumClasses());
  }

  /**
   * Returns relative frequency of class over all bags.
   */
  public final double prob(int classIndex) {
    if (!Utils.eq(totaL, 0)) {
      return m_perClass[classIndex]/totaL;
    }
    else {
      return 0;
    }
  }

  /**
   * Returns relative frequency of class for given bag.
   */
  public final double prob(int classIndex,int intIndex) {

    if (Utils.gr(m_perBag[intIndex],0))
      return m_perClassPerBag[intIndex][classIndex]/m_perBag[intIndex];
    else
      return prob(classIndex);
  }

  /**
   * Subtracts the given distribution from this one. The results
   * has only one bag.
   */
  public final Distribution subtract(Distribution toSubstract) {

    Distribution newDist = new Distribution(1,m_perClass.length);

    newDist.m_perBag[0] = totaL-toSubstract.totaL;
    newDist.totaL = newDist.m_perBag[0];
    for (int i = 0; i < m_perClass.length; i++) {
      newDist.m_perClassPerBag[0][i] = m_perClass[i] - toSubstract.m_perClass[i];
      newDist.m_perClass[i] = newDist.m_perClassPerBag[0][i];
    }
    return newDist;
  }

  /**
   * Returns total number of (possibly fractional) instances.
   */
  public final double total() {

    return totaL;
  }

  /**
   * Shifts given instance from one bag to another one.
   *
   * @exception Exception if something goes wrong
   */
  public final void shift(int from,int to,Instance instance)
       throws Exception {

    int classIndex;
    double weight;

    classIndex = (int)instance.classValue();
    weight = instance.weight();
    m_perClassPerBag[from][classIndex] -= weight;
    m_perClassPerBag[to][classIndex] += weight;
    m_perBag[from] -= weight;
    m_perBag[to] += weight;
  }

  /**
   * Shifts all instances in given range from one bag to another one.
   *
   * @exception Exception if something goes wrong
   */
  public final void shiftRange(int from,int to,Instances source,
			       int startIndex,int lastPlusOne)
       throws Exception {

    int classIndex;
    double weight;
    Instance instance;
    int i;

    for (i = startIndex; i < lastPlusOne; i++) {
      instance = (Instance) source.instance(i);
      classIndex = (int)instance.classValue();
      weight = instance.weight();
      m_perClassPerBag[from][classIndex] -= weight;
      m_perClassPerBag[to][classIndex] += weight;
      m_perBag[from] -= weight;
      m_perBag[to] += weight;
    }
  }
}